Hi Wim

I can prepare a trimmed down version of my project plus the cyana  
input and output files for you. The swapping information is buried  
within the log produced by cyana although there is probably some way  
to get this written to a separate file.

As Brian and Rasmus rightly said I will have to validate the  
stereospecific assignments cyana makes by some other method before  
including it in my project.

I assume this would generally be the case so a wholesale  
stereospecific assignment by format converter using the cyana  
assignment would definitely not be as useful as an option that would  
allow the user to define which diastereotopic assignments to include  
and which to keep ambiguous,

cheers

Marie


Quoting eiso <[log in to unmask]>:

> Wim,
>
> CANDID (or actually the noeasign macro) produces this kind output during
> the process of deciding the stereospecific assignments (so one would have
> to save the output to get this information, or compare the ppm values before
> and after this step, since it results in swapping the ppm values for the
> prochiral pairs.
>
>     Atoms with consistent swapping in 20 or more structures:
>                                 TF gap   # 1   5   10   15   20
>    2  PRO    HG2   HG3      4.9356  20 ******************** swapped
>    3  LEU    HB2   HB3      0.5458  20 ******************** swapped
>    4  GLY    HA2   HA3      2.9233  20 -------------------- as input
>    7  ASP    HB2   HB3      3.3372  20 -------------------- as input
>    9  PRO    HB2   HB3      1.7833  20 -------------------- as input
>    9  PRO    HD2   HD3      1.6049  20 -------------------- as input
>   10  ILE    HG12  HG13     5.8636  20 -------------------- as input
>   11  PRO    HB2   HB3      6.2975  20 -------------------- as input
>   11  PRO    HG2   HG3      2.5352  20 ******************** swapped
>   11  PRO    HD2   HD3      4.7207  20 ******************** swapped
>   13  GLU    HG2   HG3      3.2729  20 -------------------- as input
>
>
> then it also produces a file which defines which stereo assignments
> should be fixed.
>
> %> head  finalstereo.cya
> [...]
> atom stereo "HG2  HG3     2"   # PRO
> atom stereo "HB2  HB3     3"   # LEU
> atom stereo "HA2  HA3     4"   # GLY
> atom stereo "HB2  HB3     7"   # ASP
> atom stereo "HB2  HB3     9"   # PRO
> atom stereo "HD2  HD3     9"   # PRO
> atom stereo "HG12 HG13   10"   # ILE
>
>
> Would it be an option to check on import of cyana .prot files
> whether a swap minizes the ppm difference between the cyana
> assignments and the analysis assignments, and swap them in
> the analysis project, under certain circumstances
> for example:
> - if they are not already fixed in analysis, swap the shifts
>   in analysis to match/follow the cyana assignments
> - if they are already stereospecifically assigned within analysis,
>   the cyana shifts or resonances should/could be swapped to match the
>   analysis shifts, or maybe require some user input
> - depending on whether you trust the CANDID stereospecific assignments
>   or not, fix the stereospecific assignments in analysis if they are
>   fixed in CANDID.
>
> I think it's good not to mix up two separate issues here, one
> whether stereospecific assignments form an(y) outside source can
> be trusted, which I guess should be ultimately a user decision,
> and should anyway be possible to revert at any later stage.
> and second whether analysis or the formatconverter should be
> able to deduce &| import stereospecific assignments from an
> outside source like CANDID, ideally yes of course, but difficult
> to get right, because there are quite a few possible scenario's
> with different finegrainedness.
>
>
> goodluck,
>
> Eiso
>
>
> Wim Vranken wrote:
>> Hi Marie,
>>
>> In general, you can't tell from the .prot file if an atom was     
>> stereospecifically assigned or not, and you can definitely not tell  
>>    from the .prot file whether originally, for Asp for example, the  
>>    HBa was written out as HB2 or HB3. I do track this information   
>> when   exporting, and try to use it again when importing (as well   
>> as  shift  information), but because swaps occur in the assignment   
>> it  is truly  a mess.
>>
>> The FormatConverter tries to deal with all this as best it can, and  
>>    does use the Analysis prochiral assignments if it can, but  
>> really    there should be a direct software link between CYANA and  
>> CCPN to    deal with this correctly.
>>
>> If there is a file that lists the CYANA swaps and/or stereospecific  
>>    assignments (compared to the original input list) I could   
>> probably   use that (if you send me a CCPN project, the CYANA input  
>>  files and   the CYANA output files for testing, along with a   
>> description with   what should be happening), but any solutions   
>> coming out of that   would not be available till the next release.
>>
>> Bye,
>>
>> Wim



Marie Phelan

Joseph Black Building
Kings Buildings
West Mains Road
Edinburgh
EH9 3JJ
0131 6504704


-- 
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.